|Publication number||US8743543 B2|
|Application number||US 13/544,839|
|Publication date||3 Jun 2014|
|Filing date||9 Jul 2012|
|Priority date||1 Jun 2006|
|Also published as||CA2653817A1, CA2653817C, CN101501599A, CN101501599B, EP2310926A1, EP2310926A4, EP2310926B1, US7738251, US8218322, US20080055846, US20100091449, US20100251629, US20130083476, WO2007139560A1|
|Publication number||13544839, 544839, US 8743543 B2, US 8743543B2, US-B2-8743543, US8743543 B2, US8743543B2|
|Inventors||Jimmy Clidaras, William Whitted, William Hamburgen, Montgomery Sykora, Winnie Leung, Gerald Aigner, Donald L. Beaty|
|Original Assignee||Google Inc.|
|Export Citation||BiBTeX, EndNote, RefMan|
|Patent Citations (126), Non-Patent Citations (40), Referenced by (6), Classifications (17), Legal Events (1)|
|External Links: USPTO, USPTO Assignment, Espacenet|
This application is a continuation of, and claims the benefit of, U.S. patent application Ser. No. 12/815,892, entitled “Modular Computing Environments,” and filed Jun. 15, 2010, now U.S. Pat. No. 8,218,322 which is a continuation of U.S. patent application Ser. No. 11/757,101, entitled “Modular Computing Environments,” and filed Jun. 1, 2007, now U.S. Pat. No. 7,738,251 which claims the benefit of prior U.S. provisional application 60/810,451, entitled “Modular Computing Environments,” and filed Jun. 1, 2006.
This description relates to modular computing environments.
Computers have become widely adopted for various applications, for both personal and corporate uses. Some computers operate as stand-alone data-processing equipment, with perhaps a peripheral such as a printer and a display device. Although useful for many purposes, additional features and benefits may be achieved if multiple computers are networked together to share information and resources.
A computer network may be formed by connecting two or more computing devices with a communication channel. One type of network is a local area network (LAN). A typical residential LAN, for example, may connect two computers to a printer. A typical corporate LAN, for example, may allow many users to share resources and large amounts of information, including databases and application software.
A second type of network is a wide area network (WAN). An example of a WAN is the Internet. WANs such as the Internet allow many computer devices to communicate messages and share information. When multiple LANs are interoperable with one or more WANs, opportunities for computing devices to communicate and share information greatly expand.
From the perspective of an individual computing device that is connected to a network, users may direct the communication of information over a network with a user interface generated by a web browser application. A web browser is typically configured to enable the user to access web sites on the Internet or the World Wide Web. Web browsers allow users to easily send and receive messages over a network in packets of information. Such packets of information may include the address of a search engine website, such as www.dogpile.com, for example.
The popularity and simplicity of sharing information over networks, such as the Internet, has resulted in demand for data processing and/or storage capacity to support high network traffic volume. One mechanism to address this need may be referred to as a data center. In the context of the Internet, a data center may provide one or more of processing, storage, and support functions that improve performance or enhance the utility of the Internet. Data centers may also be deployed in other contexts. Financial institutions, for example, may employ one or more data centers to store financial account and transaction information.
A data center may provide data processing and/or storage capacity. In operation, a data center may be connected to a network, and may receive and respond to various requests from the network to retrieve, process, and/or store data. In addition to extensive data processing and data storage capabilities, data centers typically support high speed data transfer and routing capabilities. To meet future network demands, data center capacity may continue to expand.
A computer system (e.g., a data center) may include a connecting hub having a plurality of docking regions and be configured to provide to each docking region electrical power, a data network interface, a cooling fluid supply and a cooling fluid return; and a plurality of shipping containers that each encloses a modular computing environment that adds a predetermined or default amount of computing power to the system. Each shipping container may include a) a plurality of processing units (e.g., servers) coupled to the data network interface, each of which include a microprocessor; b) a heat exchanger configured to remove heat generated by the plurality of processing units by circulating cooling fluid from the supply through the heat exchanger and discharging it into the return; and c) docking members configured to releaseably couple to the connecting hub at one of the docking regions to receive electrical power, connect to the data network interface, and receive and discharge cooling fluid.
In some implementations, each processing unit may further include at least one storage device. The docking members may include quick-connect docking members. The connecting hub may include a spine; the connecting hub may include a ring.
In some implementations, a modular data center includes a connecting hub and a plurality of containers. The connecting hub may have a plurality of docking regions and may be configured to provide to each docking region electrical power, a data network interface, a cooling fluid supply and a cooling fluid return. Each container may enclose a modular computing environment having a plurality of computing devices that collectively add a large, predetermined amount of computing power to the modular data center. Each container may further have a) a first heat exchange circuit configured to transfer heat from the computing devices to a heat exchanger; b) a second heat exchange circuit comprising the heat exchanger, the cooling fluid supply and cooling fluid return, the second heat exchange circuit being configured to transfer heat from the heat exchanger inside of the container to a system outside of the container via cooling fluid in the cooling fluid supply and cooling fluid return; and c) docking members that releaseably connect to the spine at one of the plurality of docking regions to receive electrical power, connect to the data network interface, receive cooling fluid from the cooling fluid supply, and discharge return cooling fluid to the cooling fluid return.
In some implementations, each modular computing environment includes a) a plurality of microprocessor boards that receive power from the connecting hub and receive data from a network coupled to the data network interface, process the received data and transmit the processed data to the network via the data network interface; b) an air recirculating system that draws air from a cold air plenum adjacent to the plurality of racks, across the plurality of microprocessor boards, and into a warm air plenum. The heat exchanger may cool air from the warm air plenum and discharge the cooled air into the cold air plenum.
In some implementations, the modular computing environment may be configured to accommodate human occupancy. The modular computing environment may include a fire and smoke detection system and a fire suppression system. The modular computing environment may include an internal walkway and a means for ingress and egress at one or both ends of the walkway. The internal walkway may be disposed in the cold air plenum.
In some implementations, the air recirculating system may include a first set of fans and a second set of fans. Each fan in the first set may be disposed on one of the plurality of microprocessor boards, and each fan in the second set may be disposed adjacent to the heat exchanger. Adjacent fans in the second set may be supplied by electrical power from diverse electrical circuits.
In some implementations, the modular data center further includes a duct disposed between the heat exchanger and at least a first fan and a second fan in the second set. The duct may fluidly couple a region adjacent to the first fan and a region adjacent to the second fan. The plurality of microprocessor boards may be disposed in racks that are suspended from a ceiling of a corresponding container.
In some implementations, at least one of the plurality of containers may be stacked on top of another one of the plurality of containers. The modular data center may further include a facility-level cooling system that is disposed in the second heat exchange circuit, coupled to the cooling fluid supply and cooling fluid return. The facility-level cooling system may be configured to remove heat from cooling fluid in the cooling fluid return.
In some implementations, each container is sealed against environmental elements of wind and moisture. The modular data center may further include a normally sealed drain configured to open when in contact with an appreciable amount of liquid. In some implementations, at least two of the plurality of containers may be standard size shipping containers. The standard size may be a 1AAA shipping container. The standard size may be a 1CC shipping container.
In some implementations, a method of deploying a modular data center includes a) providing a cooling system configured to provide cooling fluid to various cooling devices, receive return cooling fluid from the various cooling devices and remove heat from the received return cooling fluid; b) providing an electrical power source; c) providing a data communication interface; d) routing electrical power from the electrical power source, a connection to the data communication interface, and a cooling fluid supply and cooling fluid return from the cooling system to a plurality of docking regions that each have stubs that are configured to couple the electrical power, cooling fluid supply, cooling fluid return and data communication interface connection to a modular computing environment; e) providing modular computing environments at least two of the plurality of docking regions, each modular computing environment configured to incrementally add computing power to a data center and including a plurality of computing resources and integrated cooling devices; and f) coupling each self contained computing environment at a corresponding docking region to the stubs to provide electrical power, cooling fluid supply and cooling fluid return, and connection to the data communication interface to the modular computing environment.
In some implementations, a method of processing data includes transmitting a request for data to a system and receiving data in response to the request. The system may include a) a connecting hub having a plurality of docking regions and configured to provide to each docking region electrical power, a data network interface, a cooling fluid supply and a cooling fluid return; and b) a plurality of shipping containers. Each shipping container may enclose a modular computing environment that adds a replacement amount of computing power to the computer system. Each shipping container may also have 1) a plurality of processing units that are coupled to the data network interface; 2) a heat exchanger configured to remove heat generated by the plurality of processing units from the shipping container by circulating cooling fluid from the cooling fluid supply through the heat exchanger and discharging heated cooling fluid into the cooling fluid return; and 3) docking members configured to releaseably connect to the connecting hub at one of the plurality of docking regions to receive electrical power, connect to the data network interface, receive cooling fluid from the cooling fluid supply, and discharge return cooling fluid to the cooling fluid return.
The details of one or more implementations are set forth in the accompanying drawings and the description below. Other features and advantages will be apparent from the description and drawings, and from the claims.
Like reference symbols in the various drawings indicate like elements.
As one example application of a data center or other massively parallel computer system, a financial institution may employ a number of data centers at different physical locations to store customers' financial account and transaction information or to process various transactions. When a customer of the financial institution makes a withdrawal or deposit from a teller station, a computer programming running at the teller station may receive input from the teller identifying the customer's account and specifying the type and amount of the transaction; the computer program may then send information to one or more data centers to cause the customer's account to be adjusted to reflect the withdrawal or deposit. Because the data corresponding to the customer's account may be replicated in multiple physical locations, multiple data centers may be involved in the transaction.
At substantially the same time the first customer is making a withdrawal or deposit from a teller station, a second customer may be initiating online a transfer of funds or a payment of a bill from his or her home computer. A computer program running on a web server that is accessible from the second customer's home computer may also cause the second customer's account information to be updated at multiple data centers; these data centers may be the same or different data centers than those involved in the transaction of the customer making the withdrawal or deposit from the teller station.
Many other financial transactions involving other customers of the financial institution may take place substantially simultaneously, from various computer terminals within different physical offices of the financial institution and from various other computer terminals that access the financial institution's network from the outside (e.g., home or office computer terminals of customers accessing the financial institution's services via the internet). Accordingly, the data centers may be required to process in parallel a large number of transactions in a short amount of time. Such a requirement may translate into a need for a massively parallel computing environment having a large number of processors to execute code, a large storage capacity, and a high-bandwidth network interface.
As another example application, an online information provider that supplies content and services (such as a search engine function) to many users substantially simultaneously may employ a number of data centers at different physical locations to store the supplied content and data needed to provide the search engine service or to satisfy data access latency requirements or design parameters. Deploying multiple data centers that are relatively close to large numbers of users can be particularly advantageous for delivering time-sensitive content or applications. In addition, replicating data at different data centers in different locations can be an important part of a data security strategy. In the search engine example, some of the supplied content may be physically stored by the information provider's computing system; other data that may be identified in response to a user's invocation of the search engine function may be indexed by and in the information provider's computing system. In particular, various computer programs running on the information provider's computer system may automatically navigate or “crawl” networks such as the Internet to identify and index information that has not yet been indexed or that has been updated since last being indexed. The indexing process may involve temporarily retrieving the information, parsing and analyzing it to identify a theme or topic, and storing a corresponding source identifier of the information based on the theme or topic in an information retrieval tool such as a database or hash table.
An online information provider may be required to handle a very large volume of simultaneous data traffic. For example, the information provider's computing system may receive numerous requests at approximately the same time from computer terminals of users attempting to employ the information provider to locate sources of information corresponding to particular topics. Computer programs running on the information provider's computing system may receive, parse and analyze the various requests; convert them into an appropriate format; submit them to interfaces corresponding to the information retrieval tool; receive responses from the information retrieval tool identifying content sources that store information corresponding to the requests; and transmit data associated with the content sources to the computer terminals from which the requests were originally received.
In some implementations, the transmitted data may include a pointer (e.g., a hyperlink) to the corresponding information. In some implementations, the transmitted data may include a “snippet,” or a portion, of the corresponding information (e.g., a sentence fragment from a document that includes a search term submitted with a request). In the latter case, in which a snippet is included with data transmitted to a computer terminal in response to a request received from that search terminal, snippets corresponding to various indexed information may be stored in the index or in other data storage facilities associated with the index or with the information provider.
In order to simultaneously receive and process a large number of requests from users for search engine services, the information provider may need a massively parallel computing environment having a large number of processors to execute code, a large storage capacity, and a high-bandwidth network interface.
As another example, one or more modular computing environments may be used as a high-volume “portable” data storage device. Large quantities (e.g., tens of terabytes) of frequently accessed data may be copied to one or more modular computing environments, and the modular computing environments may be transported (e.g., by cargo ship, truck or train) to a point of use. As a more concrete example, an insurance company may employ a modular computing environment to create a temporary claims office at the site of a major natural disaster. In particular, following, for example, a severe hurricane, an insurance company may store information about its policyholders in a modular computing environment and transport the modular computing environment to the site of the hurricane to create a temporary claims office. In such implementations, pre-loading large amounts of data in a modular computing environment may be more efficient or cost effective than remotely accessing the data, or transferring the data to other devices at or near the hurricane site.
As another example, one or more modular computing environments may be used as a laboratory network to simulate, in a test environment, a large network such as an enterprise intranet or a portion of the Internet. The modular computing environment may provide enough processing power and storage capacity to mimic a live network, while still remaining disconnected from live systems or live data; such an application may provide network equipment designers or manufacturers an opportunity to test or validate new hardware designs in an environment that approximates the intended operating environment without disrupting actual live data. Various components and aspects of massively parallel computing environments that may be used in the example applications outlined above are described in detail with reference to the figures.
As shown, the modular computing environment 100 includes various computing resources, an integrated cooling system, an integrated power distribution system, and an integrated networking interface. In some implementations, the modular computing environment 100 is easily transported, protected from environmental elements and thus suitable for use both inside or outside a building, configured to be easily and quickly connected with other modular computing environments to form larger computing systems, and easily serviced by human operators. For example, in some implementations, as shown, the modular computing environment is protected by an enclosure 101, such as a shipping container. The enclosure is sealed on the outside from environmental elements and equipped with access to the interior for human operators to service the various computing resources.
The computing resources of the modular computing environment 100 include a large number of processor boards. As used in this description, “processor board” refers generally to a board that includes computing resources. In some implementations, as described in greater detail with reference to
The processor boards may be physically disposed in racks, and racks carrying processor boards may substantially line the modular computing environment along the length of one or both sides of the enclosure. For example, in some implementations, the modular computing environment may include well over 1,000 processor boards. As shown in
Each processor board may be coupled to a network internal to the modular computing environment, and each processor board may further be accessible from a network external to the enclosure. In some implementations, the network internal to the modular computing environment includes a patch panel 107. The patch panel 107 may provide access connections (e.g., an RJ-45 Ethernet jack or optical interface) to network interfaces on each processor board, and the access connections may be coupled to the processor board network interfaces by, for example, category 5 or category 6 cable or fiber optic cable. The network may further include a router 110 to digitally route data from a high-bandwidth data connection external to the enclosure to processor boards within the enclosure. The network may also include one or more switches (not shown) to digitally route data from one processor board in the enclosure to another processor board in the enclosure.
A high-bandwidth connection 113 may couple the patch panel to one or more network interface connections 158 configured to connect to a network external to the modular computing environment 100. The high-bandwidth connection 113 may include one or more high-speed copper interfaces or one or more optical interfaces. For example, the high-bandwidth connection may include a gigabit Ethernet access point for balanced twisted-pair copper wire and an optical interface for connection to single- or multi-mode optical fiber. In particular, the high-bandwidth network connection may include a 10-gigabit Ethernet interface for coupling single-mode optical fiber carrying SONET (synchronous optical network signals) at carrier level 192 (OC-192) to an optical router 110.
In some implementations, a monitor/control terminal 111 may be provided to allow a human operator to monitor the various connections on the patch panel 107, switches or router 110, or to monitor actual data traffic. In some implementations, the monitor/control terminal 111 may enable a human operator to dynamically switch interconnections between various processor boards within the modular computing environment 100, or connections between an external network and various processor boards.
During operation, the computing resources may generate a significant amount of heat within the enclosure 101, and as such, the modular computing environment 100 includes an integrated cooling system to remove the generated heat from the enclosure 101. In some implementations, the integrated cooling system includes a number of heat exchangers 116 and a number of fans or pumps 119 to circulate cooling fluid (e.g., water or refrigerant) and/or fluid to be treated (e.g., air) through the heat exchangers 116. In particular, the heat exchangers 116 may be liquid-air heat exchangers that transfer heat in air that is cycled through them to a cooling fluid that is also cycled through the heat exchangers 116. The cooling fluid may be cycled through a system external to the enclosure 101, where the heat may be removed. Each heat exchanger 116 may be supplied with cooling fluid from a cooling fluid supply line 122; cooling fluid that has been warmed by the heat exchanger 116 may be discharged into a cooling fluid return line 125. Additional details of cooling systems are provided with reference to other figures.
The modular computing environment 100 also includes a power distribution system that receives power from outside the enclosure and distributes it to various loads within the enclosure. Power may be received through a connection 128 to an external power supply. Once inside the enclosure, the power may be routed to a power distribution box 131 and distributed through a number of circuits to power the computing resources, cooling system, network switching equipment and other electrical loads within the enclosure 101. The power distribution box 131 may include various circuit breakers or other over-current protection for each circuit, and each circuit may also include surge suppression devices or filters to smooth out the power or to prevent high frequency noise generated within the enclosure from being coupled into the power supply outside the enclosure. Individual circuits may be designed to minimize disruptions caused by failure of a single circuit. Additional details of an example power distribution system are provided with reference to other figures.
In order for the modular computing environment 100 to be serviced, the interior of the enclosure 101 may need to accommodate human occupancy; this may require additional features that satisfy both physical human occupancy requirements and any legal or safety requirements that may exist (e.g., municipal building or occupancy requirements). For example, the modular computing environment 100 may include interior lights, a source of fresh air, fire detection and suppression systems, provisions for safe ingress and egress of human occupants; and temperature, humidity and noise may be controlled to be within certain ranges.
Additional structural details of the example modular computing environment 100 are now described. As mentioned above, the modular computing environment 100 is protected by an enclosure 101 that is sealed from environmental elements such as, wind, rain or snow. In some implementations, the enclosure 101 is a standard shipping container. For example, the enclosure may be an ISO 1AAA container, sometimes referred to as a “high cube” 40′ container, having an approximate inside width of 7′8″, an approximate inside height of 8′10″, and an approximate inside length of 39′5″. Other container dimensions are possible. In particular, the modular computing environment 100 may be enclosed by a standard height 20′ container (e.g., an ISO 1CC container having an inside height of approximately 7′10″ and an inside length of 19′4″); or the container may be longer (e.g., 45′, 48′ or 53′). Moreover, enclosures other than shipping containers may be used. For example, in some implementations, the enclosure may be a metal- or wood-framed enclosure with protective siding material. Racks within the enclosure 101 may be suspended (e.g., from the ceiling or wall, to provide additional space under the racks for running cables, installing other equipment (e.g., cooling equipment) or providing addition service access), and the walkway 142 may be segmented into small, removable sections to facilitate access to the components under the racks (e.g., heat exchangers 116 or fans 119).
The enclosure 101 may be configured to be easily transportable. Shipping containers are particularly well-suited for this purpose and are designed to be easily transported, for example, by a cargo ship, train or truck. The enclosure may also be stackable and may include locking mechanisms (not shown) to secure multiple enclosures in a stacked arrangement (e.g., in operation). The enclosure may include fork-lift pockets 143 and corner fittings 146 to facilitate handling by cargo transport devices (e.g., fork-trucks, cranes, overhead booms, conveyors, etc.).
In some implementations, a modular computing environment 100 may be transported to a desired location and immediately deployed by being connected to an electrical power supply, a cooling system and a network interface, as described in greater detail below. In some implementations, a modular computing environment 100 may be deployed while it is still disposed on a transport device. For example, a modular computing environment 100 may be transported to an intended deployment site by cargo ship, train or truck, and the modular computing environment may be connected to electrical power a cooling system and a network while it is still on the cargo ship, train or truck. In general, the modular computing environment 100 may be configured to be rapidly deployed.
The modular computing environment 100 includes ports that facilitate quick connection to utilities and other services required for operation. For example, as shown in
Modular computing environments 100 with protective enclosures 101 as described above may have several benefits. For example, such modular computing environments 100 may facilitate quick assembly of large data centers. Large portions of a data center may be prefabricated and quickly deployed; in particular, portions of data centers may be constructed in parallel, rather than in sequence. Critical portions of data centers may be mobile and easily transported from one site to another. Portions of the data center may be manufactured by manufacturing labor, rather than constructed by trade labor (e.g., in a controlled manufacturing environment rather than in an uncontrolled construction site), possibly resulting in reduced costs.
To move air through the heat exchanger 116 from an input region (corresponding to path A) to an output region (corresponding to path C), the cooling system 201 includes multiple fans 119. The fans 119 may be mounted to a duct 202 that fluidly couples a space in front of the heat exchanger 116. The duct 202 may help to balance variations in pressure along the duct's length; in addition, the duct 202 may improve the overall reliability of the cooling system 201 by pulling a large volume of air through the heat exchanger 116, even if one of the fans 119 stops functioning. For example, even if a middle fan 119 stops functioning, the duct 202 may even out pressure variations relative to the outside fans 119 that may otherwise exist and pull air through the middle of the heat exchanger 116. As shown, the fans 119 pull air through the heat exchanger 116, but in other implementations, fans may be disposed such that they push air through the heat exchanger 116.
An overall cooling system may include other portions, in addition to the portion 201 that is illustrated in
As shown, the modular data center 400 includes two modular computing environments 100A and 100B. (The modular data center 400 may include many other modular computing environments, but only two are shown for purposes of illustration.) The modular computing environments 100A and 100B include a number of computing devices 405 that generate heat within the enclosure 100A. Air may be circulated past the computing devices 405 to absorb the generated heat (e.g., via path D). The modular computing environment also includes one or more integrated heat exchangers 116, through which the air may be circulated to remove the absorbed heat (e.g., via path B). In some implementations, as described above, the heat exchanger 116 is a liquid-air heat exchanger in which heat removed from the air is transferred to a cooling fluid circulated through the heat exchanger 116 (e.g., via supply and return lines 122 and 125).
As shown, the transfer of heat from the computing devices 405 to air and from air to cooling fluid comprises a first thermal circuit 401. In some implementations, the amount of heat generated by the computing devices 405 is substantially equivalent to the amount of heat transferred to the cooling fluid (ignoring heat received or dissipated through walls of the enclosure 101A). In these implementations, substantially all of the generated heat may be removed by a second thermal circuit 402.
As shown, the second thermal circuit 402 includes the cooling fluid supply line 122, a cooling fluid supply main 412 that provides cooling fluid to multiple modular computing environments 100A and 100B, the cooling fluid return line 125, a cooling fluid return main 415 that receives cooling fluid from multiple modular computing environments 100A and 100B, and a cooling system 420 (e.g., a refrigeration-based chiller or condenser, an evaporative cooling tower, or another heat exchanger or heat pump). In operation, the second thermal circuit 402 may function at a facility level (e.g., the entire modular data center 400) to exhaust heat from multiple modules (e.g., the modular computing environments 100A and 100B). Other types and arrangements of the cooling system 420 are possible.
In some implementations, heat may flow in other ways. For example, some heat may be exhausted from the enclosure 101 corresponding to the modular computing environment 100 by advection, e.g., through a fresh-air ventilation system to an external environment 518. In many implementations, such heat transfer has only second-order effects relative to the primary transfer of heat to the second cooling fluid 512; that is, the transfer of heat from the heat-generating circuits 501 to the external environment 518 by first cooling fluid 506 that leaves the enclosure may be negligible vis-à-vis the transfer of heat from the heat-generating circuits to the second cooling fluid 512. As another example, some heat may be transferred to the body 515 of the enclosure 101 by convection, and then transferred from the body 515 of the enclosure 101 by conduction or radiation. In many implementations, this form of heat transfer also only has second-order effects relative to the primary transfer of heat to the second cooling fluid 512; that is, the transfer of heat from the heat-generating circuits 501 to the external environment 518 by radiation from the body 515 of the enclosure may be negligible vis-à-vis the transfer of heat from the heat-generating circuits to the second cooling fluid 512.
For simplicity, the description above accounts for heat transfer out of the enclosure 101. However, in some implementations, heat may also enter the enclosure 101 from the external environment. For example, heat may be transferred by convection or radiation to the body 515 of the enclosure 101, and from the body 515 of the enclosure 101 to the first cooling fluid 506. In many implementations, such heat transfer also has only second-order effects, and a detailed description is therefore omitted.
In some implementations, a cable 609 (e.g., a Category 5, Category 6, or optical fiber cable) may couple a network interface on a processor board to the patch panel 107 (see
In some implementations, the power cable 702 is connected to a circuit that is split out in the power distribution box 131 (e.g., a 277 VAC, single-phase circuit). In some implementations, power is distributed from the power distribution box 131 to various rack supports along the top of the rack; in other implementations, power may be distributed in a different manner (e.g., via buses, along the bottom of the rack, under the walkway, to the front of the racks etc.)
In operation, the processor board 800 may execute computer program instructions that are stored in the storage devices 808A or 808B or memory 805, or that are received via the network interface 814. In multi-processor implementations, each processor 802A or 802B may independently execute instructions, and each processor 802A or 802B may have corresponding dedicated memory or storage devices. The instructions may, for example, cause the processors to retrieve data from the storage devices 802A or 802B and provide it to another device coupled to the network interface 814, process data, or store data in the storage devices 802A or 802B or memory 805. A large data center may employ thousands of processor boards such as the processor board 800 to perform complex data manipulations, such as, for example, managing customer account information within a financial institution, or providing various users with content and information services.
Electrical power may be provided to the modular data center 900 in the form of medium voltage power (e.g., 13 kV) by a power utility via utility lines 916. The voltage of the incoming power may be reduced by a facility-level transformer 919 (e.g., to 480 VAC) and distributed at a facility level power distribution box 922 and routed to the various modular computing environments within the spine 913. In some implementations, other facility backup or supplemental power may be provided. For example, a diesel backup generator (not shown) may be co-located with the modular data center 900; if the utility power is compromised, the diesel backup generator may provide interim power to the modular data center.
A central cooling plant 925 may be provided at the facility level to provide cooling fluid to each modular computing environment 901-910 for the removal of heat generated within corresponding enclosures of the modular computing environments. As shown (and described above with reference to
A central communication facility 928 may also be provided at the facility level and may couple the network interfaces of each modular computing environment 901-910 to a larger network. An example communication facility 900 is graphically depicted as a satellite dish, but the communication facility could be any kind of high-bandwidth network connection (e.g., OC-192, 10-gigabit Ethernet, parallel T1 or T3 lines, etc.) and associated networking equipment (e.g., switches, routers, gateways, etc.). In some implementations, the communication facility 928 couples the modular data center 900 to a company's intranet (e.g., the internal network of a financial institution or an Internet information provider); in some implementations, the communication facility 928 couples the modular data center 928 to the Internet.
In some implementations, the spine 913 provides a layer of protection to the utilities (e.g., electrical power, cooling fluid and a network interface) delivered to each modular computing environment 901-910. The utilities may be delivered via flexible “quick-connection” stubs disposed at “docking regions” (not shown) corresponding to locations along the spine 913 that are configured to receive modular computing environments. For example, electrical power may be delivered from a distribution line or bus (not shown) within the spine 913 to heavy-duty “appliance style” connectors that can be plugged into a modular computing environment. Similarly, cooling fluid may be delivered from large supply and return mains (not shown in
Deploying the modular data center 900 may be a simple process. In some implementations, the process includes a) providing (e.g., constructing) a facility level cooling plant 925; b) providing a facility-level electrical power system 916, 919 and 922; c) providing a facility-level communication facility 928; d) routing cooling fluid, electrical power and a connection to the facility-level communication interface to various docking regions (not shown), which may be optionally located in a spine 913; e) providing two or more modular computing environments 901-910 at two or more of the docking regions; and f) connecting the utilities to the modular computing environments using the quick-connection stubs disposed at the docking regions.
The example method of deploying the modular data center may have one or more of the following advantages: modular data centers may be quickly deployed; a modular data center 900 may be constructed in parallel (e.g., modular computing environments may be constructed in parallel with facility level resources); the modular data center 900 may be easily expanded to incrementally add computing power; large portions of the modular data center 900 may be mobile to facilitate relocation; and portions of the modular data center 900 may be prefabricated, possibly at lower cost than would be possible if the modular data center 900 were serially constructed from scratch (e.g., some manufacturing labor may be used in place of trade labor). Moreover, portions of a modular data center may be manufactured in a controlled factory environment, rather than at a less controlled construction site, thereby reducing overall costs associated with lower quality control.
Other implementations are contemplated. For example, although ten modular computing environments 901-910 are shown, a data center may have many more (e.g., 10-100 modular computing environments); the modular computing environments 901-910 could have other arrangements, such as a star configuration centered around utility and network distribution points within a ring-shaped connecting hub (not shown), or a stacked configuration coupled to a vertical spine as illustrated in
In some implementations, each modular computing environment 1106A or 1106B includes a phase splitter 1112 that splits off each of the three phases to single-phase legs having a voltage V3. The phases may be split in a manner that balances the overall load of each modular computing environment. In some implementations, the voltage V3 is single-phase 277 VAC. The three single-phase legs at voltage V3 may be distributed as various circuits to loads within the modular computing environment 1106A or 1106B. For example, networking switching equipment 1115 may have a dedicated circuit. A fan, or a number of fans 1118A and 1118B may have a dedicated circuit (in some implementations, adjacent fans may be supplied by different circuits for reliability). Lights 1121 may have a dedicated circuit. Groups of processing boards 1124A-1124C may have dedicated circuits (e.g., a column of processor boards in a rack, or another portion of a rack 1125A, 1125B or 1125C may have a dedicated circuit). Other devices 1127 may also have dedicated circuits (e.g., lighting, fire/smoke detection devices and a fire suppression system). In some implementations, circuits are designed to be small enough that disruption of a circuit (or a small fraction of the total number of circuits) will not critically disrupt operation of the modular computing environment. In particular, for example, cooling components can be powered by multiple circuits so that disruption of one circuit will not significantly impact the overall cooling capacity of the modular computing environment. As shown, each circuit at the voltage V3 may have over-current devices 1130 (e.g., circuit breakers, fuses, surge suppressors, etc.). In some implementations, each processor board 1124A-C includes an on-board power supply 1133 that converts electrical power having voltage V3 to one or more voltages suitable for components on the corresponding processor board.
Other implementations are contemplated. For example, at the facility level, additional power sources, such as backup or supplemental generators may be provided. Single-phase power may be used throughout a facility in some implementations. Voltages other than those described above may be employed.
Various components of the example information provider 1201, as shown, are now described. The information provider 1201 may include content stores 1213A and 1213B, which may store specific kinds of information, such as, for example, maps of various locations, or images related to particular topics. The information provider 1201 also includes information modules 1216A and 1216B, which may each provide specific services such as determining driving directions between two locations, or providing timely weather information for a specific location. The information provider 1201 includes a search engine 1219 that may be configured to locate content that is internal or external to the information provider 1201 (e.g., content in content stores 1213A or 1213B, or content stored by the content providers 1210A-C).
To provide a search engine function, the information provider 1201 may maintain an index 1222 (e.g., an index database) and may further use system storage 1225 to maintain the index or provide the search engine functions. The index database 1222 may contain data that represents information that the information provider 1201 provides to users. For example, the index database 1222 may contain links to information stored on content providers 1210A-C outside of the information provider 1201. Some information referenced by entries in the index database 1222 may be stored in the system storage 1225. For example, the system storage 1225 may “mirror” information for which search reports are regularly received—such as, for example, breaking news stories or weather or traffic information. The system storage 1225 may also store various components needed for general operation of the information provider 1222, such as, applications, system parameters, and information about users who access the system.
The index database 1222 may or may not be cached. For example, the index database 1222 may include a separate cached index database (not shown) to support faster access to search results. The system storage 1225 may be local to the information provider 1225, or it may be distributed, such as, for example, in an external server or storage farm (not shown). Within the information provider 1201, the search engine 1219 may operate to produce search results that include information stored in the index database 1222 or the system storage 1225, in response to search requests from users.
As shown, the information provider 1201 receives and transmits data through an external interface 1228 and an internal interface 1231. Through the external interface 1228, which may include one or more Web servers (not shown), the information provider 1201 receives requests and transmits responses from and to external computing devices (e.g., the client device 1204). Through the internal interface 1231, the information provider 1201 internally routes information to and from the index database 1222 and the system storage 1225 via various paths (some of which are not shown), and to and from the information modules 1216A and 1216B and the content stores 1213A and 1213B.
Each interface 1228 and 1231 may include one or more distinct interfaces. For example, a first interface may receive requests, such as, for example, requests for information or search queries; a second interface may transmit results to requesting devices. These interfaces may comprise interface devices for a high-speed, high-bandwidth network such as SONET, Infiniband, or Ethernet, or any suitable communication hardware operating under an appropriate protocol such that the information provider 1201 can respond to a large number of distinct requests simultaneously. The precise design of the interfaces 1228 and 1231 or of the overall information provider 1201 is not critical to this description and may take any suitable form.
Information in the index database 1222 may be gathered by an automated information gatherer 1234, such as, for example, a Web crawler or a spider. The automated information gatherer may, for example, continuously, or almost continuously, obtain new information from sources connected to the network 1207 or to other networks. The automated information gatherer 1234 may also retrieve content in response to a search query. This content may be provided to the index database 1222 or to the system storage 1225, or to both. The index database 1222 may index data that is accessible from the network 1207 or from another network.
In addition to being added to the index database 1222 or system storage 1225 in an automated fashion, information may also be manually loaded in or retrieved from the index database 1222 or the system storage 1225 through a maintenance interface 1237. For example, a maintenance interface 1237 may allow an administrator of the information provider 1201 to manually add bulk data to the index database 1222 or to the system storage 1225.
Data requests, such as queries received from users, may be received and processed through a request processor 1240. The request processor 1240 may, for example, parse requests, and, if necessary, reformat them (e.g., from HTML (hypertext markup language) or text format, to search term or strings that are compatible with the search engine 1219). The request processor 1240 may operate in conjunction with the external interface 1228. In some embodiments, the request processor 1240 may be included in the external interface 1228.
Responses to data requests may be formatted for delivery by a response formatter 1243. The response formatter 1243 may, for example, format content that is responsive to data requests in a format like HTML, XML (Extensible Markup Language), WML (Wireless Markup Language), or some other suitable format. The response formatter 1243 may operate in conjunction with the external interface 1228. In some embodiments, the response formatter 1243 may be included in the external interface 1228.
The information provider 1201 may further include an ad delivery system 1246, along with an ad database 1249 and an ad log 1252. The ad delivery system 1246 may receive query requests and other input from the internal interface 1231. In some embodiments, the ad delivery system 1246 may select an ad from the ad database 1249 to deliver in conjunction with other content, such as, for example, a response to the query. The selected ad may correspond to content of the search query. For example, in response to a user who submits a query for a restaurant near a particular location having a particular cuisine (e.g., That food), the information provider may deliver, along with information about various restaurants having that cuisine in that location, a coupon for a particular restaurant. The response formatter 1243 may combine search results from the search engine 1219 and ad results from the ad delivery system 1246 into an overall response and may send that response to the device from which the original query was receive, via the network 1207. An ad maintenance module 1254 may be provided to add, manage and remove ads in the ad database 1249
Some ads may solicit a response from the recipient. For example, an ad may include a link that a user can select to receive additional information or to perform other actions. Any such responses may be directed back to the information provider 1201 and subsequently logged. For example, the information provider 1201 may receive, via the interfaces 1228 and 1231 and ad delivery system 1246, a response to a previously delivered ad. The ad delivery system 1246 may store this response in the ad log 1252. The ad log 1252 may be subsequently used to bill sponsors of such ads or to log various statistics related to ads and to their effectiveness.
A number of embodiments have been described. Nevertheless, it will be understood that various modifications may be made without departing from the spirit and scope of the disclosed embodiments. Accordingly, other embodiments are within the scope of the following claims:
|Cited Patent||Filing date||Publication date||Applicant||Title|
|US4407185||23 Feb 1982||4 Oct 1983||Borg-Warner Corporation||Return air flow control for variable air volume system|
|US4589712||17 May 1984||20 May 1986||Hastings Otis||Data processing equipment enclosures|
|US5119014||5 Mar 1991||2 Jun 1992||Kronberg James W||Sequential power-up circuit|
|US5317859||30 Sep 1992||7 Jun 1994||Advanced Pulver Systems, Inc.||Product orienter and loader|
|US5458991||19 May 1993||17 Oct 1995||Sl Waber, Inc.||UPS with auto self test|
|US5565274||8 Sep 1995||15 Oct 1996||The Bilco Company||Fire rated floor door and control system|
|US5579522||24 Jul 1995||26 Nov 1996||Intel Corporation||Dynamic non-volatile memory update in a computer system|
|US5595068||15 Dec 1995||21 Jan 1997||Carrier Corporation||Ceiling mounted indoor unit for an air conditioning system|
|US5596844||3 Feb 1995||28 Jan 1997||Kalinowski; Juan R.||Foldable portable building|
|US5694307||30 Sep 1996||2 Dec 1997||Alliedsignal Inc.||Integrated AC/DC and DC/DC converter|
|US5722886||29 Feb 1996||3 Mar 1998||Steckly; Stephen Dale||Ventilator|
|US5746950||18 Dec 1996||5 May 1998||Villanueva; Robert||Drain hose assembly|
|US5844884||23 Apr 1997||1 Dec 1998||Nokia Mobile Phones Limited||Battery monitoring|
|US6009177||19 Feb 1997||28 Dec 1999||Certco Llc||Enhanced cryptographic system and method with key escrow feature|
|US6034873||2 Jun 1998||7 Mar 2000||Ericsson Inc||System and method for separating air flows in a cooling system|
|US6138413||12 Dec 1996||31 Oct 2000||Huron Window Corporation||Standardized framing section for closure wings|
|US6138935||3 Jun 1997||31 Oct 2000||Zebco Division Of Brunswick Corp||Two piece cover for fishing reel|
|US6249221||28 Jul 1999||19 Jun 2001||Joyce J. Reed||Emergency detector door illumination escape system|
|US6289684||14 Jan 2000||18 Sep 2001||Guidry, Ii John L.||Transportable, self-contained, refrigeration system|
|US6290024||3 Mar 2000||18 Sep 2001||Lubrication Systems Company Of Texas, Inc.||Oil mist generating system|
|US6374627||9 Jan 2001||23 Apr 2002||Donald J. Schumacher||Data center cooling system|
|US6429706||24 Oct 2001||6 Aug 2002||Juniper Networks, Inc.||Voltage sequencing circuit for powering-up sensitive electrical components|
|US6496366||7 Jan 2000||17 Dec 2002||Rackable Systems, Llc||High density computer equipment storage system|
|US6504717||15 Jun 2001||7 Jan 2003||Cereva Networks. Inc.||Failure-tolerant high-density card rack cooling system and method|
|US6506111||10 Aug 2001||14 Jan 2003||Sanmina-Sci Corporation||Cooling airflow distribution device|
|US6563706||23 Jun 1998||13 May 2003||Hewlett-Packard Development Company, L.P.||Typically high-availability information storage product|
|US6584559||28 Jan 2000||24 Jun 2003||Avaya Technology Corp.||Firmware download scheme for high-availability systems|
|US6587343||27 Nov 2001||1 Jul 2003||Sun Microsystems, Inc.||Water-cooled system and method for cooling electronic components|
|US6652373||16 May 2002||25 Nov 2003||Sanmina-Sci Corporation||Cooling airflow distribution device|
|US6652374||25 Nov 2002||25 Nov 2003||Sanmina-Sci Corporation||Cooling airflow distribution device|
|US6667891||20 Feb 2001||23 Dec 2003||Rackable Systems, Inc.||Computer chassis for dual offset opposing main boards|
|US6668565||12 Apr 2002||30 Dec 2003||American Power Conversion||Rack-mounted equipment cooling|
|US6672955||3 Sep 2002||6 Jan 2004||International Business Machines Corporation||Air flow management system for an internet data center|
|US6741467||23 Oct 2002||25 May 2004||Rackable Systems, Inc.||High density computer equipment storage system|
|US6747872||28 Feb 2003||8 Jun 2004||Hewlett-Packard Development Company, L.P.||Pressure control of cooling fluid within a plenum|
|US6768222||11 Jul 2000||27 Jul 2004||Advanced Micro Devices, Inc.||System and method for delaying power supply power-up|
|US6822859||24 Jun 2003||23 Nov 2004||Rackable Systems, Inc.||Computer rack cooling system|
|US6824362||15 Jul 2002||30 Nov 2004||Adc Dsl Systems, Inc.||Fan control system|
|US6850408||24 Jul 2003||1 Feb 2005||Rackable Systems, Inc.||High density computer equipment storage systems|
|US6859366||19 Mar 2003||22 Feb 2005||American Power Conversion||Data center cooling system|
|US6867967||16 Dec 2002||15 Mar 2005||International Business Machines Corporation||Method of constructing a multicomputer system|
|US6880349||29 Sep 2003||19 Apr 2005||American Power Conversion Corporation||Rack-mounted equipment cooling|
|US6896612||26 Jan 2004||24 May 2005||Sun Microsystems, Inc.||Self-cooled electronic equipment enclosure with failure tolerant cooling system and method of operation|
|US6904968||14 Sep 2001||14 Jun 2005||Hewlett-Packard Development Company, L.P.||Method and apparatus for individually cooling components of electronic systems|
|US6927980||26 Jan 2004||9 Aug 2005||Hitachi, Ltd.||Cooling structure for disk storage device|
|US6935130||24 Jun 2003||30 Aug 2005||Aopen Inc.||Computer cooling system|
|US6937947||20 Aug 2002||30 Aug 2005||Lsi Logic Corporation||Battery charger system and method for providing detailed battery status and charging method information|
|US6967283||2 Jan 2002||22 Nov 2005||American Power Conversion Corporation||Adjustable scalable rack power system and method|
|US6969922||25 Feb 2005||29 Nov 2005||Youtility, Inc||Transformerless, load adaptive speed controller|
|US6980433||23 Dec 2004||27 Dec 2005||American Power Conversion Corporation||Data center cooling system|
|US7011576||15 May 2002||14 Mar 2006||Sanmina Sci Corporation||Cooling airflow distribution device|
|US7046514||7 Jun 2004||16 May 2006||American Power Conversion Corporation||Data center cooling|
|US7106411||5 May 2004||12 Sep 2006||Imax Corporation||Conversion of cinema theatre to a super cinema theatre|
|US7123477||31 Mar 2004||17 Oct 2006||Rackable Systems, Inc.||Computer rack cooling system|
|US7145772||16 Sep 2005||5 Dec 2006||American Power Conversion Corporation||Data center cooling system|
|US7168935||1 Aug 2003||30 Jan 2007||The United States Of America As Represented By The Administrator Of The National Aeronautics And Space Administration||Solid freeform fabrication apparatus and methods|
|US7173820||14 Apr 2006||6 Feb 2007||American Power Conversion Corporation||Data center cooling|
|US7173821||17 Jun 2003||6 Feb 2007||Rackable Systems, Inc.||Computer rack with power distribution system|
|US7214843||14 Oct 2003||8 May 2007||Exxon Mobil Chemical Patents Inc.||Treating oxygenate containing feedstreams in the conversion of oxygenates to olefins|
|US7278273||30 Dec 2003||9 Oct 2007||Google Inc.||Modular data center|
|US7295442 *||14 Jun 2002||13 Nov 2007||Sun Microsystems, Inc.||System management|
|US7382613||19 May 2005||3 Jun 2008||Hewlett-Packard Development Company, L.P.||Computer system with external air mover|
|US7500911||14 Feb 2005||10 Mar 2009||American Power Conversion Corporation||Exhaust air removal system|
|US7511959||25 Apr 2007||31 Mar 2009||Hewlett-Packard Development Company, L.P.||Scalable computing apparatus|
|US7511960||13 Sep 2006||31 Mar 2009||Sun Microsystems, Inc.||Balanced chilled fluid cooling system for a data center in a shipping container|
|US7525207||26 Feb 2007||28 Apr 2009||Google Inc.||Water-based data center|
|US7529086||12 Jan 2007||5 May 2009||American Power Conversion Corporation||Data center cooling|
|US7604535||27 Apr 2006||20 Oct 2009||Wright Line, Llc||Assembly for extracting heat from a housing for electronic equipment|
|US7646590||6 Jul 2007||12 Jan 2010||Exaflop Llc||Data center power distribution|
|US7701714 *||16 Jan 2007||20 Apr 2010||Flextronics Ap, Llc||Liquid-air hybrid cooling in electronics equipment|
|US7752858||25 Nov 2002||13 Jul 2010||American Power Conversion Corporation||Exhaust air removal system|
|US7841381||22 Apr 2005||30 Nov 2010||Stirling Technology, Inc.||Heat and energy recovery ventilators and methods of use|
|US7987223 *||14 Jun 2002||26 Jul 2011||Oracle America, Inc.||Interfacing computer modules|
|US20020007643||26 Apr 2001||24 Jan 2002||Toc Technology, Llc||Computer rack heat extraction device|
|US20020055329||8 Nov 2001||9 May 2002||Storck Gary A.||Raised floor air handling unit|
|US20030002254||29 Jun 2001||2 Jan 2003||Intel Corporation||High capacity air-cooling system for electronic apparatus and associated methods|
|US20030133263||17 Jan 2002||17 Jul 2003||H.L.L. Investment, Ltd.||Uninterruptible power supply (UPS) with built-in USB hub|
|US20030150231||10 Mar 2003||14 Aug 2003||Toc Technology, Llc||Computer rack heat extraction device|
|US20030181158||30 Jan 2003||25 Sep 2003||Edwards Systems Technology, Inc.||Economizer control|
|US20030197428||3 Apr 2003||23 Oct 2003||Hatton Thomas E.||Power processor|
|US20030209023||9 Jun 2003||13 Nov 2003||Toc Technology, Llc||Computer room air flow method and apparatus|
|US20030237008||24 Jun 2002||25 Dec 2003||Compaq Information Technologies Group, L.P.||Multiple server in-rush current reduction|
|US20040020224||4 Dec 2002||5 Feb 2004||Bash Cullen E.||Cooling system with evaporators distributed in parallel|
|US20040020232||30 Jul 2003||5 Feb 2004||Hewlett-Packard Development Company, L.P.||Underfloor cable junction unit and computer center equipped with such junction units|
|US20040054780 *||16 Sep 2002||18 Mar 2004||Hewlett-Packard Company||Dynamic adaptive server provisioning for blade architectures|
|US20040090787||28 Aug 2003||13 May 2004||Color Kinetics, Inc.||Methods and systems for illuminating environments|
|US20040095249||24 Jul 2003||20 May 2004||Robert Zaccaria||Method and apparatus for the continuous performance monitoring of a lead acid battery system|
|US20040099747||25 Nov 2002||27 May 2004||Johnson Rollie R.||Exhaust air removal system|
|US20040105758||22 Sep 2003||3 Jun 2004||Ross Carl M.||Method and system for liquid leak detection, notification and removal|
|US20040150374||3 Nov 2003||5 Aug 2004||Rudy Kraus||Apparatus for providing high quality power|
|US20040184232||19 Mar 2003||23 Sep 2004||James Fink||Data center cooling system|
|US20040223300||7 Jun 2004||11 Nov 2004||Fink James R.||Data center cooling|
|US20050012395||5 Dec 2003||20 Jan 2005||Steven Eckroad||Integrated closed loop control method and apparatus for combined uninterruptible power supply and generator system|
|US20050036248||15 Aug 2003||17 Feb 2005||American Power Conversion Corporation||Uninterruptible power supply|
|US20050088201||23 Mar 2004||28 Apr 2005||Nallatech, Ltd.||Programmable power supply system|
|US20050099750||29 Sep 2004||12 May 2005||Fumikazu Takahashi||DC backup power supply system and disk array using same|
|US20050099770||23 Dec 2004||12 May 2005||James Fink||Data center cooling system|
|US20050168945||29 Dec 2004||4 Aug 2005||Giovanni Coglitore||Computer rack cooling system with variable airflow impedance|
|US20050170770||14 Feb 2005||4 Aug 2005||American Power Conversion Corporation||Exhaust air removal system|
|US20050225936||28 Mar 2003||13 Oct 2005||Tony Day||Cooling of a data centre|
|US20050259392||19 May 2005||24 Nov 2005||Vinson Wade D||Computer system with external air mover|
|US20060007653||16 Sep 2005||12 Jan 2006||James Fink||Data center cooling system|
|US20060020832||21 Jul 2004||26 Jan 2006||Giga-Byte Technology Co., Ltd.||Circuit capable of updating power supply specification of microprocessor and method thereof|
|US20060031448 *||3 Aug 2004||9 Feb 2006||International Business Machines Corp.||On demand server blades|
|US20060065000||14 Nov 2005||30 Mar 2006||Belady Christian L||Redundant upgradeable, modular data center cooling apparatus|
|US20060082263||14 Oct 2005||20 Apr 2006||American Power Conversion Corporation||Mobile data center|
|US20060139877||29 Dec 2004||29 Jun 2006||Mark Germagian||Rack height cooling|
|US20060187636||14 Apr 2006||24 Aug 2006||Fink James R||Data center cooling|
|US20070076373||4 Dec 2006||5 Apr 2007||James Fink||Data center cooling system|
|US20070097636||4 Aug 2006||3 May 2007||Johnson Douglas A||Adaptive cooling method for computer rack enclosure|
|US20070139883 *||15 Dec 2005||21 Jun 2007||Pinkerton Joseph F Iii||Systems and methods for providing resources such as cooling and secondary power to electronics in a data center|
|US20070254583||27 Apr 2006||1 Nov 2007||Mark Germagian||Assembly for extracting heat from a housing for electronic equipment|
|US20070274035||12 Jan 2007||29 Nov 2007||Fink James R||Data center cooling|
|US20080062647||13 Sep 2006||13 Mar 2008||Sun Microsystems, Inc.||Balanced chilled fluid cooling system for a data center in a shipping container|
|US20080064317||13 Sep 2006||13 Mar 2008||Sun Microsystems, Inc.||Cooling method for a data center in a shipping container|
|US20080094797||25 Sep 2007||24 Apr 2008||Giovanni Coglitore||Container-based data center|
|US20080123288||13 Sep 2006||29 May 2008||Sun Microsystems, Inc.||Operation ready transportable data center in a shipping container|
|US20090050591||23 Jul 2008||26 Feb 2009||David Michael Hart||Mobile Data Center Unit|
|US20090195977||20 Mar 2009||6 Aug 2009||American Power Conversion Corporation||Data center cooling|
|US20090241578||31 Mar 2008||1 Oct 2009||Exaflop Llc||Warm Floor Data Center|
|USD473225||20 Feb 2001||15 Apr 2003||Rackable Systems, Inc.||Portion of a bank of computer chassis mounted to rack bars|
|USD475705||18 Feb 2000||10 Jun 2003||Rackable Systems, Inc.||Computer chassis for dual, opposing main boards|
|CN1732727A||24 Nov 2003||8 Feb 2006||美国能量变换公司||Exhaust air removal system|
|CN2687732Y||9 Apr 2004||23 Mar 2005||昆明钢铁股份有限公司||Computer casing|
|EP621451A3||Title not available|
|WO2007082351A1||18 Jan 2007||26 Jul 2007||Datatainer Pty Ltd||Data processing apparatus|
|1||"Protect High-Density Equipment From Thermal Damage-Guaranteed," APC Currents, 2007, 18(4), 3 pages.|
|2||"Protect High-Density Equipment From Thermal Damage—Guaranteed," APC Currents, 2007, 18(4), 3 pages.|
|3||‘Circuit Converts PWM Fan Drive to Linear and Reduces Acoustic Noise’ [online]. Maxim/Dallas Semiconductor, 2006, [retrieved on Oct. 23, 2006]. Retrieved from the Internet: <URL: www.maxim-ic.com/appnotes.cfm/an—pk/3530>. 3 pages.|
|4||‘Intelligent Temperature Monitor and PWM Fan Controller’ [online] Analog Devices, 2003, [retrieved on Oct. 24, 2006]. Retrieved from the Internet: <URL: www.analog.com/UploadedFiles/Data—Sheets/ADM1030.pdf#xml=http://search.analog.com/search/pdfPainter.aspx?url=http://www.analog.com/UploadedFiles/Data—Sheets/ADM1030.pdf&fterm=Intelligent&fterm=Temperature&fterm=Intelligent Temperature&la=en>. 28 pages.|
|5||‘LM64—+− 1C Remote Diode Temperature Sensor with PWM Fan Control and 5 GPIO's’ [online]. National Semiconductor, 2006, [retrieved on Oct. 23, 2006]. Retrieved from the Internet: <URL: http://www.elecdesign.com/Articles/Index.cfm?ArticleID=6281>. 3 pages.|
|6||‘Low-cost fan manager for power-supply applications’ [online]. Electronic Design, 2006, [retrieved on Oct. 23, 2006]. <URL: www.elecdesign.com/Articles/Index.cfm?ArticleID=6281>, 3 pages.|
|7||‘T10 Project 1142D—Working Draft SCSI Parallel Interface-2’ [online]. American National Standard, 1998, [retrieved on Oct. 23, 2006]. Retrieved from the Internet: <URL: www.t10.org/ftp/t10/drafts/spi2/spi2r20b.pdf >, pp. 181-182.|
|8||'Circuit Converts PWM Fan Drive to Linear and Reduces Acoustic Noise' [online]. Maxim/Dallas Semiconductor, 2006, [retrieved on Oct. 23, 2006]. Retrieved from the Internet: . 3 pages.|
|9||Cringley, Robert, "Google-Mart: Sam Walton Taught Google More About How to Dominate the Internet Than Microsoft Ever Did," I, Cringely-The Pulpit, Nov. 17, 2005, 3 pages.|
|10||Cringley, Robert, "Google-Mart: Sam Walton Taught Google More About How to Dominate the Internet Than Microsoft Ever Did," I, Cringely—The Pulpit, Nov. 17, 2005, 3 pages.|
|11||Extended European Search Report mailed dated Mar. 4, 2013 for international application No. 06785739.1 (9 pages).|
|12||Greg Matter Weblog, The Industrial Revolution, Finally, available at: http://blogs.sun.com/Gregp/date/200610, Oct. 17, 2006, 5 pages.|
|13||Hamilton, "Commodity Data Center Design" Microsoft Corp, Apr. 17, 2007, 11 pages.|
|14||Hamilton, James, "An Architecture for Modular Data Centers" Microsoft Corp,CIDR 2007. 3rd Biennial Conference on Innovative Data Systems Research (CIDR) Jan. 7-10, 2007, Asilomar, California, USA pp. 306-313.|
|15||'Intelligent Temperature Monitor and PWM Fan Controller' [online] Analog Devices, 2003, [retrieved on Oct. 24, 2006]. Retrieved from the Internet: <URL: www.analog.com/UploadedFiles/Data-Sheets/ADM1030.pdf#xml=http://search.analog.com/search/pdfPainter.aspx?url=http://www.analog.com/UploadedFiles/Data-Sheets/ADM1030.pdf&fterm=Intelligent&fterm=Temperature&fterm=Intelligent Temperature&la=en>. 28 pages.|
|16||International Preliminary Report on Patentability in International Application No. PCT/US2006/024845, mailed Dec. 18, 2008, 7 pages.|
|17||International Preliminary Report on Patentability in International Application No. PCT/US2006/24844, mailed Dec. 18, 2008, 6 pages.|
|18||International Search Report and Written Opinion in International Application No. PCT/US2006/024845, mailed Dec. 18, 2006, 9 pages.|
|19||International Search Report and Written Opinion in International Application No. PCT/US2006/024880, mailed Feb. 15, 2008, 14 pages.|
|20||International Search Report and Written Opinion in International Application No. PCT/US2006/039715, mailed Nov. 5, 2007, 12 pages.|
|21||International Search Report and Written Opinion in International Application No. PCT/US2006/039928, report completed Mar. 25, 2007, 8 pages.|
|22||International Search Report and Written Opinion in International Application No. PCT/US2006/24844, mailed Apr. 26, 2007, 10 pages.|
|23||International Search Report and Written Opinion in International Application No. PCT/US2006/25163, mailed May 3, 2007, 8 pages.|
|24||'LM64-+- 1C Remote Diode Temperature Sensor with PWM Fan Control and 5 GPIO's' [online]. National Semiconductor, 2006, [retrieved on Oct. 23, 2006]. Retrieved from the Internet: . 3 pages.|
|25||'Low-cost fan manager for power-supply applications' [online]. Electronic Design, 2006, [retrieved on Oct. 23, 2006]. , 3 pages.|
|26||Office Action mailed Dec. 9, 2010 for Chinese application No. 200680055450.4 (12 pages).|
|27||Rackable Systems, "Rackable Systems Contcentro Modular Data Center High Density, High Efficiency Mobile Computing," Datasheet, 2007, 1 page.|
|28||Rackable Systems, "Thermal Management & Server Density: Critical Issues for Today's Data Center," White paper, 2004, 8 pages.|
|29||Sanmina-SCI Product Sheet, ECOBAY, "High-Performance Closed-Loop System," 2006, 2 pages.|
|30||Shankland, Stephen, "Sun Blackbox, meet APC's whitebox," cNet News.com, available at http://news.com.com/Sun+Blackbox%2C+meet+APCs+whitebox/2001-1010-3-617774.html, Jan. 7, 2007, 2 pages.|
|31||Shankland, Stephen, "Sun Blackbox, meet APC's whitebox," cNet News.com, available at http://news.com.com/Sun+Blackbox%2C+meet+APCs+whitebox/2001-1010—3-617774.html, Jan. 7, 2007, 2 pages.|
|32||Sun Project Blackbox-Index, Available at: http://www.sun.com/emrkt/blackbox/index.jsp, Retrieved Oct. 19, 2006, 3 pages.|
|33||Sun Project Blackbox—Index, Available at: http://www.sun.com/emrkt/blackbox/index.jsp, Retrieved Oct. 19, 2006, 3 pages.|
|34||Sun Project Blackbox-Scenarios, Available at: http://www.sun.com/emrkt/blackbox/scenarios.jsp, Retrieved Oct. 19, 2006, 7 pages.|
|35||Sun Project Blackbox—Scenarios, Available at: http://www.sun.com/emrkt/blackbox/scenarios.jsp, Retrieved Oct. 19, 2006, 7 pages.|
|36||Supplementary European Search Report in European Application No. EP 0678558.2-1242, dated Jun. 30, 2010, 6 pages.|
|37||Supplementary European Search Report mailed Jun. 30, 2010 for international application No. 06785588.2, PCT/U52006/024845 (6 pages).|
|38||System Management Interface Forum, "PMBus Power System Management Protocol Specification Part 1-General Requirements, Transport and Electrical Interface, Revision 1.0" Mar. 2005, 17 pages.|
|39||System Management Interface Forum, "PMBus Power System Management Protocol Specification Part 1—General Requirements, Transport and Electrical Interface, Revision 1.0" Mar. 2005, 17 pages.|
|40||'T10 Project 1142D-Working Draft SCSI Parallel Interface-2' [online]. American National Standard, 1998, [retrieved on Oct. 23, 2006]. Retrieved from the Internet: , pp. 181-182.|
|Citing Patent||Filing date||Publication date||Applicant||Title|
|US9069533 *||15 Nov 2012||30 Jun 2015||Hong Fu Jin Precision Industry (Shenzhen) Co., Ltd.||Data center container with draining mechanism|
|US9609784 *||18 Mar 2013||28 Mar 2017||Hon Hai Precision Industry Co., Ltd.||Container data center|
|US20130229769 *||15 Nov 2012||5 Sep 2013||Xu Yang||Data center container with draining mechanism|
|US20140199934 *||18 Mar 2013||17 Jul 2014||Hon Hai Precision Industry Co., Ltd.||Container data center|
|US20150163963 *||9 Dec 2014||11 Jun 2015||Naver Business Platform Corporation||Booth apparatus for supplying cooling air|
|US20160338229 *||1 Mar 2016||17 Nov 2016||Microsoft Technology Licensing, Llc||Modularization of data center functions|
|U.S. Classification||361/701, 709/249, 454/118, 165/80.4, 62/259.2|
|International Classification||G06F15/16, B60F1/00, F28F7/00, G06F13/00, F25D23/12, H05K7/20|
|Cooperative Classification||G06F2200/201, H05K7/1497, H05K7/202, G06F1/20, H05K7/2079, H05K7/20754|
|16 Jan 2013||AS||Assignment|
Owner name: GOOGLE INC., CALIFORNIA
Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:CLIDARAS, JIMMY;WHITTED, WILLIAM;HAMBURGEN, WILLIAM;AND OTHERS;SIGNING DATES FROM 20070807 TO 20071101;REEL/FRAME:029644/0839